Readmission to the hospital: common, complex and time for a re-think

Healthcare costs and demand for services continue to soar internationally. In a previous era of fee-for-service reimbursement, healthcare utilisation was less problematic from a hospital’s perspective because it was a source of revenue. Hospitals and payers now scrutinise healthcare utilisation, including hospital readmissions, due to the shift towards value-based reimbursement models. However, the metric and methods we use to evaluate progress in managing hospital readmissions are problematic. When payers and health services researchers operationalise readmission as a binary outcome—yes or no—they vastly oversimplify the actual phenomenon. This truncation of meaning has unintended consequences when payers use binary readmission measures in value-based purchasing programmes, and when scientists use them in research.

Across the world, healthcare systems have implemented guidelines and incentives to decrease hospital readmissions with some success. Yet questions have been posed as to whether large-scale improvements are the result of real change or other games in the system. Clinicians have attributed pejorative terms like ‘frequent flyer’ to patients, framing readmission as a behavioural problem. Although conceptually alluring and useful, risk prediction models for readmission have shown inconsistent performance among individuals with high-risk chronic conditions like heart failure and kidney disease (Alba et al., 2013).

We believe there is a misalignment between readmission as it is conceptualised, how clinicians and organisations respond to a value-based agenda, and the way patients and caregivers experience illness. We will explore this misalignment by describing the system, clinician and patient factors associated with hospital readmissions through discussing the Hospital Readmission Reduction Program. Although the discussion of the Hospital Readmission Reduction Program provided below is directly related to the United States (US) healthcare system, the policy critique has international relevance, particularly as healthcare systems across the world grapple with readmissions in the context of population ageing and the increasing burden of chronic illness. The US, Australia, Canada, Denmark, England and Germany have developed their own readmission measures. The US, England and Germany have implemented financial incentives to reduce readmissions, and Australia is currently exploring the same (Australian Commission on Safety & Quality in Health Care, 2019).

1 |. THE US HOSPITAL READMISSION REDUCTION PROGRAM AND SUBSEQUENT HEALTHCARE SYSTEM CHANGES

In its 2007 report to the US Congress, the Medicare Payment Advisory Commission concluded that ‘potentially preventable’ hospital readmissions cost Medicare billions of dollars each year. The Commission also identified ‘target conditions’ with presumably modifiable variation in 30-day readmission rates, including heart failure, pneumonia and acute myocardial infarction (Medicare Payment Advisory Commission, 2007). In 2012, the Centers for Medicare and Medicaid Services (CMS) implemented the Hospital Readmission Reduction Program to reduce this variation and incentivise high quality inpatient care (Zuckerman et al., 2016). Researchers created binary readmission measures for target conditions with rigorous and transparent methods (Yale New Haven Health Services Corporation - Center for Outcomes Research & Evaluation, 2019). Centers for Medicare and Medicaid Services can now reduce Medicare payments to hospitals based on their historical 30-day readmission rates for target conditions compared to the national average (Zuckerman et.al., 2016).

When CMS created the Hospital Readmission Reduction Program, there was little available evidence to guide the system redesign that would lower readmission rates. In its 2007 report, the Medicare Payment Advisory Commission acknowledged that readmission reductions would require intervention at multiple healthcare settings across the care continuum and patient engagement in care (Medicare Payment Advisory Commission, 2007). Over the last 10 years, readmission rates have declined (Wasfy et al., 2017) Interventions focusing on care coordination and patient centredness have shown promise (Hansen et al., 2011). However, healthcare leaders in the US have raised concerns about the utility and safety of the Hospital Readmission Reduction Program. In one retrospective cohort study, Hospital Readmission Reduction Program implementation was associated with an increase in 30-day post-discharge mortality for patients with heart failure and pneumonia (Wadhera et al., 2018). Others have suggested that nationwide reductions in readmission rates may be attributable to increases in emergency room utilisation, hospital observation stays (Nuckols et al., 2018) and discharges against medical advice (Onukwugha & Alfandre, 2019); signals that patients with complex medical conditions may be receiving inadequate care. Responding to scepticism, in 2017 Krumholz and colleagues demonstrated that the same patients had different hospital-wide readmission rates after discharge from bottom and top performing hospitals (Krumholz et al., 2017). Though their research strengthened the validity of the hospital-wide readmission measure, the ongoing debate among clinicians and researchers indicates that key issues are unresolved.

2 |. CLINICIAN AND HOSPITAL RESPONSES TO VALUE-BASED PURCHASING PROGRAMS

A second key issue with binary hospital readmission measures and value-based purchasing programmes like the Hospital Readmission Reduction Program is scepticism about whether punitive financial incentives motivate clinicians and the hospitals they work in to improve patient outcomes by changing the structure or process of care (Burgess & Hockenberry, 2014; Woolhandler et al., 2012). Existing quality improvement theory and empirical evidence support this scepticism. In stark contrast to the premises of value-based payment, W. Edwards Deming, an icon in quality improvement, classified pay for performance as a ‘force of destruction’ that ‘robs the company and the nation of innovation and applied science’ and ‘crowds out intrinsic motivation, self-esteem, dignity and joy in work’(Edwards Deming, 2013). Decades of research in psychology on motivation demonstrate that extrinsic motivation (e.g. incentives or rewards) works best when incentives are directly related to performance on concrete tasks that are relatively simple (Cerasoli et al., 2014). In contrast, value-based purchasing programmes like the Hospital Readmission Reduction Program penalise (not reward) hospitals (not individuals) for retrospective, pooled outcomes that reflect highly complex and interdependent organisational, clinician and patient behaviours.

Value-based purchasing programmes, like the Hospital Readmission Reduction Program, may also contribute to clinician dissatisfaction and frustration. As previously mentioned, observation stay rates in the US have increased over the last decade as readmission rates have declined. Though research refutes the link between the Hospital Readmission Reduction Program and an increase in observation stays (Zuckerman et al., 2016), a recent Society of Hospital Medicine white paper stated, ‘The observation issue…can severely damage the therapeutic bond with patient/family who may conclude that the hospitalist has more interest in saving someone money at the expense of the patient’ (Society of Hospital Medicine, 2017). Readmission regulation and value-based purchasing programmes are complex. Clinicians may remain sceptical, creating a chain of distrust between patient, clinician and hospital administration at a time when burned out care teams need human connection the most (Cochrane et al., 2019).

3 |. PATIENT AND CAREGIVER EXPERIENCES OF READMISSION

Lastly, existing hospital readmission measures and value-based purchasing programmes to reduce readmissions may not be patient-centred (Umscheid & Greysen, 2018). Binary readmission measures impose a negative value on readmission that does not reflect patient or caregiver experience of readmission as a rational choice. For example, heart failure became a target condition in the Hospital Readmission Reduction Program because of readmission rate variation across hospitals and their relative high cost to Medicare. To hospitals that participate in the Hospital Readmission Reduction Program, a 30-day readmission for heart failure is always ‘bad’. However, patients and caregivers may hate the hospital but value the immediacy of care, symptom relief and respite. In one qualitative study, interviews of patients with heart failure and their caregivers depicted readmission as a rational choice (not ‘good’ or ‘bad’)—one that patients and caregivers made together as they adapted to new constraints and symptoms of disease, often through trial and error (Sevilla-Cazes et al., 2018). When a funding model penalises hospitals for care that its beneficiaries value, it may do patients, caregivers and communities a disservice.

A more concerning consequence of imposing a negative value on readmission is the allocation of blame. A recent study demonstrated that providers were more likely to identify patient factors as contributing to readmission, whereas patients were more likely to identify system factors, for example the need for earlier follow-up with a doctor. Nurse case managers evaluated each readmission case and, in alignment with patient perspectives, identified system factors most of the time (Smeraglio et al., 2019). When providers and other members of the care team are ‘dinged’ for patient outcomes, they may be more likely to assign blame for readmission to patients.

This may be especially true for low income or otherwise margin-alised patients living in under-resourced communities (McHugh et al., 2010). Authors of one qualitative study of readmission for heart failure distinguished between immediate, precipitating and underlying factors associated with readmission. For example, one participant who relied on food stamps and free meals at churches was readmitted for shortness of breath (immediate factor) related to dietary non-adherence (precipitating factor) caused by homelessness and food insecurity (underlying factor) (Sentell et al., 2016). Patients and caregivers navigating underlying socioeconomic factors may rationally leverage hospital resources when needed. Hospital readmission, then, may be a symptom of distributive injustice rather than a pathology all its own. In their compelling case–control study of low-and high-performing hospitals in the Hospital Readmission Reduction Program, Caracciolo and colleagues found that poverty, food stamp participation rates and other socioeconomic indicators in the surrounding community were associated with hospital penalty and questioned ‘whether poor performance on quality measures is a function of underperforming hospitals or a manifestation of underserved communities’ (Caracciolo et al., 2017).

4 |. QUESTIONS TO FRAME THE WAY FORWARD

Binary readmission measures are not working for key stakeholders in the healthcare system. They have unintended consequences for patients. There have been significant advances in quality measurement, reporting, and infrastructure; but a lack of conceptual alignment around ‘what quality is’ and how to incentivise it has created tensions for key stakeholders across the healthcare system. Resolving those pain points will require collaboration from each of them. The following ‘wicked questions’ may help frame key issues:

• How can we leverage forward-thinking quality frameworks that remind us we are not giving quality care if it burns out our workforce?

• How do we create quality measures that reflect deeper patient values—not simply outcome variation—without a full reset on entrenched payment models?

• What non-punitive incentives can we use to change organisational and clinician behaviour?

• How can the value of nursing care be demonstrated in outcome measures

Through auditing programmes to identify and recover improper payments to healthcare providers in the US, there have been significant resources put into developing criteria for ‘appropriate readmissions’ and the infrastructure to audit billing. Perhaps it is time to develop ‘appropriateness criteria’ for readmissions that (a) are person-centred and (b) incentivise removal of system barriers in the healthcare system. We have the data reporting and auditing oversight to add nuance to measurement and avoid unintended consequences of a binary outcome in which readmissions are always bad regardless of circumstance.

5 |. CONCLUSION

Commonly in health care, we use jargon and terms, which are highly value laden. Readmission has taken on this quality. Terming the use of appropriate healthcare utilisation in the context of chronic and complex disease is a crucial issue for concept development and testing in prospective models in order to inform evidence-based policy decisions.